Study Of Risk Prediction After Acute Thrombectomy By The Secondary Cerebral Collateral Circulation Partitioning Combined With A Multimodal Score

Purpose This study aims to analyse the diagnostic efficacy of meticulous partitioning combined with computed tomography（CT）multimodal scoring of the secondary collateral circulation pial artery in predicting hemorrhagic transformation（HT）after mechanical thrombectomy in acute ischemic stroke（AIS）.Methods According to the Alberta Stroke Program Early CT Score（ASPECTS）,we divided the secondary pial collateral circulation in the blood supply range of the middle cerebral artery into five regions,namely the M2 or M6 regions of ASPECTS.On the basis of multiphase CT angiography（m CTA）,we evaluated the pial arterial filling status of every region in extent,delay,and contrast washout.We established the prediction models of HT and symptomatic intracerebral hemorrhage（s ICH）by using CT perfusion（CTP,model-H1 and model-S1）,m CTA（model-H2 and model-S2）and comprehensive parameters（model-H3 and model-S3）.We used the receiver operating characteristic curve to analyse the prediction performance of each model.The 90-day modified Rankin Scale（90d-m RS）score was used as the evaluation standard for mid-term clinical outcomes of AIS patients.We compared the differences in the mid-term clinical function recovery outcome of the AIS patients between asymptomatic intracerebral hemorrhage（a ICH）and s ICH,as well as whether or not combined with intravenous thrombolysis before mechanical thrombectomy.Results Among the 112 patients with AIS who received thrombectomy,39（34.8%）developed HT,17（15.2%）of whom had s ICH and the other 22（19.6%）had a ICH.In model-H1 and model-S1,relative surface permeability was an independent risk factor for HT（OR,95%CI:1.119,1.020–1.229）and s ICH（OR,95%CI:1.081,1.002–1.165）after thrombectomy,and the areas under the curves（AUCs）of diagnostic models were 0.928 and 0.929,respectively.In model-H2 and model-S2,washout was an independent protective factor for HT（OR,95%CI:0.615,0.439–0.860）and s ICH（OR,95%CI: 0.543,0.356–0.827）,and the AUCs were 0.920 and 0.952,respectively.Model-H3 and model-S3 improved the prediction performance（AUC: HT,0.939;s ICH,0.964）.Significant differences were observed in the 90d-m RS scores of the AIS patients between non-HT,a ICH,and s ICH groups（P<0.001）.The 90d-m RS scores of the s ICH group were significantly higher than those of the non-HT group（P<0.001）and a ICH group（P=0.009）,but no significant difference was noted between the latter two groups（P>0.05）.The proportion of mid-term neurologic independence（90d-m RS ≤2）in the three groups was 58.2%,33.3%,and 7.1%,respectively,,and the differences were significant（P=0.001）.No significant differences were observed in the 90d-m RS score and the mid-term independent ratio of neurological function between mechanical thrombectomy combined with intravenous thrombolysis and independent mechanical thrombectomy group（P>0.05）.Moderate and weak correlations were found between the parameters of CTP and the three filling statuses of the pial artery.The relative cerebral blood volume（r CBV）was the only CTP parameter that was significantly correlated with the filling status of the pial artery（P<0.05）.Conclusion Meticulous partition combined with a CT multimodal score of the secondary collateral pial artery has good application value for the risk evaluation of HT and s ICH in patients with AIS after mechanical thrombectomy.The comprehensive parameter model can improve diagnostic performance.Intravenous thrombolysis before endovascular treatment was an independent risk factor for HT after reperfusion but did not significantly affect the prognosis of AIS patients.